Eur Rasplr J
1992, 5, 1231-1238
Clara cell protein in serum and bronchoalveolar lavage
A. Bernard*, F.X. Marchandise**, S. Depelchin**, R. Lauwerys*, Y. Sibille**
Clara cell protein in serum and bronchoalveolar lavag e. A. Bernard, F.X.
Marchandise, S. Depelchin, R. Lauwerys, Y. Sibille.
ABSTRACT: The 10 kDa Clara cell protein was measured in serum and
bronchoalveolar lavage (BAL) from 39 healthy subjects (14 smokers, 25 nonsmokers) and from 41 patients with respiratory disease (chronic obstructive pulmonary
disease (COPD), sarcoidosis, lung cancer).
Clara cell protein appears as one of the most abundant respiratory tract derived
proteins, with values averaging 7% of the total protein content of lung lavages
from healthy nonsmokers. A significant reduction of Clara cell protein was found
in BAL from smokers and patients with COPD or lung cancer. The same pattern
of change was found in the concentrations of Clara cell protein in serum.
Pulmonary sarcoidosis did not affect absolute values of Clara cell protein in lung
lavages but was associated with elevated levels in serum. Changes in lung lavage
Clara cell protein differed from that of albumin, ~ 2 -microglobulin or the secretory component, since the latter were unaffected by smoking or COPD but increased in sarcoidosis and lung cancer
These results indicate that Clara cell protein in BAL or serum might serve as
a sensitive indicator of nonciliated bronchial cell dysfunction.
Eur Respir J., 1992, 5, 1231-1238.
The bronchoalveolar lavage (BAL) procedure allows
cells and soluble components to be collected from distal
parts of the respiratory tract. In healthy subjects, 6080% of proteins in BAL derive from serum by passive
transudation [1]. The remainder comprises proteins
secreted locally, in particular by epithelial cells. Among
these epithelial cell derived products, the secretory component, recognized as a transepithelial carrier of polymeric immunoglobulins A and M (IgA and IgM), can
represent up to 10% of the BAL protein content [2, 3].
Recently, a 10 kDa protein has been described in the
lung of rat, rabbit and man [4-7]. This small protein is
the major secretory protein of the Clara cell, hence it is
named Clara cell protein (CC10). We have demonstrated that CC10 is identical to an alpha-microprotein
(protein 1), isolated from the urine of patients with renal
tubular dysfunction [8-10]. CC10 (or protein 1) could
act as a natural immunosuppressor protecting the respiratory and urogenital tracts from exaggerated inflammatory reactions [6, 8, 10].
Using a sensitive immunoassay developed for protein 1 [11], and hence CC10, we have determined the
concentrations of CC10 in BAL and serum from
healthy subjects, and have studied the influence of
smoking and of various respiratory diseases on this
lung secretory protein.
• Industrial Toxicology and Occupational
Medicine Unit and • • Pulmonary Section
and Experimental Medicine Unit (ICP),
Catholic University of Louvain, Brussels,
Belgium.
Correspondence: A. Bernard
Unite de Toxicologie Industrielle et
Medecine du Travail
Universite Catholique de Louvain
Clos Chapelle-aux-Champs Bte 30.54
1200 Brussels
Belgium.
Keywords: Bronchoalveolar lavage
~ 2 -microglobulin
clara cell protein
lung cancer
protein
sarcoidosis
smoking
Received: April 23 1992
Accepted after revision August 4 1992
Subjects and methods
Study population
Healthy nonsmokers. Twenty five healthy volunteers, who had never smoked (17 males and 8 females), with a mean age of 28 yrs (range 21-65 yrs),
were included in the study. Their lung function tests
and lung X-rays performed just before bronchoscopy
were normal. There was no suspicion of recent viral
or bacterial infection.
Normal smokers. Fourteen healthy smokers, with a
mean age of 29 yrs (range 19-45 yrs) and with a
mean smoking history estimated at 10.4 pack-years
(range 3-20 pack-years), constituted the smoker group.
Lung function tests and lung X-rays performed just
before bronchoscopy were normal.
Chronic obstructive pulmonary disease (COPD)
patients. This group included six patients (all males),
with a mean age of 55 yrs (range 43-70 yrs).
They were all current smokers, with a mean smoking
history of 35 pack-years (range 15-55 packyears). Their forced expiratory volume in one second
1232
A. BERNARD ET AL.
(FEV1) was 1.7:t0.3 l, i.e. 55:7% of predicted values.
Their pulmonary diffusing capacity for c.arbon monoxide (DLco) and diffusion index (DLco/VA) measured
by the Cothas simple breath method (Mynhardt
transfer system) were 76:tl1% and 84:tl7% of predicted values, respectively, suggesting mild emphysema. They had Jung X-rays without signs of acute
lesions.
Sarcoidosis patients. Seventeen patients (10 men and
7 women, all lifetime nonsmokers), with biopsy proven
sarcoidosis were also included in the study. Their age
averaged 47 yrs (ranged 25-74 yrs). Six patients had
enlarged hilar and/or mediastinal lymph nodes (Stage
I), 7 combined enlarged thoracic lymph nodes and
parenchyma! lung infiltrates (Stage II) and 4 presented
with isolated lung infiltrates (Stage III).
Cancer patients. Eighteen primary lung cancer patients
(16 men and 2 women) were included in the study.
BAL samples from both the affected and unaffected
side were available for 13 patients, from the affected
side only for 2 patients, and from the unaffected side
only for 3. The diagnosis of bronchogenic carcinoma,
suspected at the time of bronchoscopy, was confirmed
on the basis of the histopathological examination
of tissues, obtained either at bronchoscopy or by
transthoracic needle lung biopsy. Fourteen patients
were smokers, with a mean smoking history evaluated at 50 pack-years (range 35-85 pack-years) and
4 were nonsmokers. Their age was 62 yrs on
average (ranged 44-79 yrs). The mean of the
Karnosfky scale was over 90 and the mean of weight
loss for the last 6 months was 3%. There were 6
squamous lung cancer, 7 adenocarcinomas, 2 limited
and 3 extended anaplastic lung cancer. Among
the nqn-anaplastic lung cancer, 6 were at stage 1
of the TNM classification (12] and 7 at Stage 3.
Table 1. -
Their FEV1 was 2.4:0.4 l, i.e. 75:6% of predicted values and their DLco and DLco/VA were 75:11% and
78:11% of predicted values respectively, (13].
BAL procedure
All patients and volunteers underwent the same
standard BAL procedure (with 250 ml of sterile saline
solution) through the fibreoptic bronchoscope as
described previously [14]. The recovery volumes,
as well as the values of some cell and protein constituents of the collected BAL fluids, are listed in
table 1.
Assay of CCJO and other proteins
The concentration of CC10 in BAL fluids was
determined by a sensitive immunoassay relying on the
agglutination of latex particles. A detailed description
of this immunoassay has been published recently in its
application to urinary protein 1 (11]. The accuracy of
CClO assay was tested by adding purified CClO to 10
samples of serum (final concentration of 100 p.g·J·1) or
BAL (final concentration of 5 mg·/' 1). The analytical
recovery (measured over a period of 2- 3 days) averaged 99% (so 10.5%) and 94% (so 9.3%), respectively.
We also assessed the stability of CC10 in 10 BAL
samples kept at 37°C for 24 h. The concentration of
CClO in these samples averaged 101% (so 12%) of
that in samples kept frozen. In Ouchterlony immunodiffusion analysis, CC10 from different BAL fluids
showed a complete identity with the protein purified
from tubular proteinuria (fig. 1). Pooled specimens
(n=S) of BAL from healthy subjects and patients were
fractionated by fast protein liquid chromotography
(FPLC) on Sephacryl S-200 (Pharmacia-LKB Biotechnology, Uppsala, Sweden) and CC10 was measured
in the eluted fractions by latex immunoassay (11 ].
Protein and cell components of bronchoalveolar lavage
Normal
nonsmokers
n
Volume recovered m!
Total cell count xlO)·mP
Macrophage count %
Lymphocyte count %
PMN count %
Albumin mg·/'t
(32-microglobulin ""g·J'l
Secretory component
mg·f'l
25
127±17
165:!:44
90.3±6.2
9.1±4.2
0.5±0.8
19.1
(5.3-59)
20
(0.9-129)
8.1
(2.7-27)
Normal
smokers
14
126:!:26
577±400•
96.2±4.7
3.1:3.6*
0.8±0.3
19.1
(5.8-64)
20.4
(0.9-83)
8.4
(1.3-23)
COPD
6
102±32
444:!:281*
97.1:1:2.1
1.1±0.2*
2.2±1.2*
12.1
(5.7-25)
9.0
(2.1-43)
11.4
(4.4-35.4)
Sarcoidosis
17
126±19
337±135*
82.7±9.9
15.2±9.3*
1.8±1.8
50.8
(7.8-328)•
107
(18- 545)•
20.6
(9.1- 94)*
Cancer patients
n=18
unaffected side
affected side
16
94±40•
390:354"
95.3±4.4
3.6±4.0*
0.9:!:1.1
13.7
(5- 35.9)
51
(0.9-820)*
15.4
(3.1-40.2)•
15
93±23*
417±208*
95.1±5.1
3.0:!:3.3*
1.6:!:2.5
19.8
(4.6-257)
30
(0.9-260)
10.7
(1.8-61.1)
Data are presented as mean±so, or geometric mean and range in brackets. • : significantly different from nonsmokers. In cancer patients, albumin values were missing for three samples (one in the affected side and two in the unaffected side). PMN:
polymorphonuclear neutrophils; COPD: chronic obstructive pulmonary disease.
1233
CLARA CELL PROTEIN IN SERUM AND BAL
In all cases, CC10 eluted as a single component with
an apparent molecular size around 16,000, which
was indistinguishable from that of the native protein
(fig. 2). It should be noted that the size of CC10 (10
kDa) is underestimated by sodium dodecyl sulphatepolycrylamide gel electrophoresis (SDS-PAGE) due
to anomalous migration properties of the intact
protein. CC10 is indeed composed of two identical
subunits, with a size between 7 and 8 kDa as estimated
by SDS-PAGE and from the amino acid composition
[10]. The fact that, by gel filtration, the protein
shows an apparent size around 16 kDa is therefore not
surprising.
Albumin and secretory component in BAL fluids
were determined by an immunoradiometric assay [14],
whereas P2-microglobulin was measured by latex immunoassay [15]. Total protein concentration in BAL
was measured by the Bio-Rad protein assay using
bovine serum albumin standard (Bio-Rad, Richmond,
California, USA).
1
:.•
6
~ 2)·
5
3
Statistical analysis
Fig. 1. - Ouchterlony immunodiffusion analysis of purified Clara
cell protein (CC10 or protein 1) (well 1); urine from a patient with
acute tubular necrosis (well 2); and bronchial lavage (BAL) fluids
from a patient with lung infection (well 3), lung cancer (well 4), sarcoidosis (well 5) and healthy smoker (well 6). The central well contained 20 !AI of anti-CClO antibody and oth er wells 20 ~-t1 (wells 3
and 4) or 40 t-tl (wells 2, 5 and 6) of solutions with CC10 concentratious between 10 and 50 mg·/· 1• After 24 h diffusion, the plate
was washed, dried and stained with Coomassie brilliant blue.
2000
0
0
All statistical tests were performed using the Slatview SE software [16]. Differences between groups
were assessed by one-way analysis of variance followed
by the Fisher's least significant difference (FLSD)
multiple comparison test, with p<0.05 considered as
significant. In patients with lung cancer, comparison
between the affected and non-affected side of the lung
was made using the paired t-test. All BAL or serum
analytes were log-transformed before statistical analysis.
0
0
0
,.._
(")
1.0
~
~
~
80
0
,.._
0
"<T
CO
~
0
N
~
i::n
:::!..
c:
0
~
c 1000
Cl)
u
c
0
u
0
G
(.)
0~--~~~~~~~--~~~~~---
10
20
30
40
50
Tube number
Fig. 2. - Distribution of Clara cell protein (CCl 0) in the fractions obtained by chromatography on Sephacryl S-200 of a pooled specimen of
bronchoalveolar lavage (BAL) fluids from healthy subjects or patients. Arrows indicate markers of relative molecular mass. e : nonsmokers;
0: smokers; A. : sarcoidosis;
lung cancer; • : purified CC10.
0:
1234
A. BERNARD ET AL.
100
A
...
.....0,
0
10
E
362!
.....J
<3::
CO
.s
c:
~
•
•
8
!*
1.7 ,.
8
8
•
0
Cl.
Q)
u
~
3.5!
•
•
0.99.!!!... *
(.)
8
••
*•
.!2
8
t
2.69 ~
t
...
...
i
1.45
-t *
§
...t
t
t
0.1 ~----.-------r-----~.------.-------r------.----
Nonsmokers
Smokers
COPD
Sarcoidosis Cancer(·)
Cancer(+)
Fig. 3. - Effect of smoking, COPD, sarcoidosis and lung cancer on Clara cell protein levels in BAL (log scale). Cancer(·): unaffected side; Cancer
(+): affected side. Closed symbols correspond to smokers and bars represent the geometric means. *: significantly different from nonsmokers; §:
significantly different from values in the unaffected side (paired t-test, n=l3). BAL: bronchoalveolar lavage; COPD: chronic obstructive pulmonary
disease.
Results
CClO represented on average 7.2:5% (so) (n=24) of
the total protein content of BAL (the latter averaged
60:30 mg·l'1). This distribution was shifted to significantly lower values in smokers, in patients with
COPD and in the affected side of patients with lung
cancer. The concentration of CClO in BAL from patients with sarcoidosis, or in the unaffected side of
patients with lung cancer, did not differ significantly
The concentration of CClO in lung lavages from
nonsmokers showed a log-normal distribution around
a mean of 3.6 mg·['l (median 4.1 mg·/' 1) (fig. 3).
The concentration of CClO in BAL did not differ
signific-antly between women (n=8, geometric mean 3.3
mg·/'1) and men (n=17, geometric mean 3.8 mg·/'1).
10
...
.....J
<3::
tO
.s
<>
0
~
c:
.E
::>
..0
~ 0.1
(.)
~
..:
0.19' ..
0
00
0
(.)
•
0
0
0.09-*
*•
0
•
•
o.os -·-
..
•
...
Nonsmokers
Smokers
~
0.07 0 " *
00
00
0
i
t
...
...
0.15 -tt"''f
...
...
...
0
•
0.01
0
0
0
"
. .. . * §
o.os+
...
...
......
CO
COPD
Sarcoidosis Cancer(·)
""'
Cancer(+)
Fig. 4. - Effect of smoking, COPD, sarcoidosis and lung cancer on the Clara cell protein/albumin ratio in BAL (log scale). *: significantly
different from nonsmokers; §: significantly different from values in the unaffected side (paires t-test, n=ll). Albumin values were missing for
three samples from cancer patients (one in the affected and two in the unaffected side). CClO: Clara cell protein. For definition of symbols
and further abbreviations see legend to figure 3.
1235
CLARA CELL PROTEIN IN SERUM AND BAL
from that observed in nonsmokers (fig. 3). Notably, the
four cancer patients who had never smoked had on average a higher CClO level in BAL (unaffected side)
than those who were smokers (6.71 vs 1.87 mg·L-1,
p=0.06). This pattern of changes of CClO in lung
lav-ages was clearly distinct from that observed with
albumin, ~ 2 -microglobulin or secretory component
(table 1).
When values of CClO in BAL were expressed relative to that of albumin (fig. 4), a significant reduction
was still present in smokers and in the affected side
of patients with lung cancer. The reduction in COPD
patients was no longer statistically significant, whereas,
by contrast, the values in sarcoidosis patients were
significantly lowered, because of the elevation of
albumin in the BAL of these patients.
1000
0
0
0
-.:...
a,
:::!.
8
E
&
2
en
Q)
100
82
.s:
0
.,....
+ +
:
8
<>
<>
(.)
(.)
134
:•
•
•
*
~
0
•
63.6
<>
<>
_Q_
0
29
...
...
......
/:;
50 -x-
...
-*
•
•••
...
10
Nonsmokers
Smokers
COPD
Sarcoidosis
Cancer
Fig. 5. - Clara cell protein (CC10) in the serum of nonsmokers and patients with COPD, sarcoidosis or lung cancer (log scale). Closed
symbols correspond to smokers and bars represent the geometric means. • significantly different from nonsmokers. COPD: chronic obstructive pulmonary disease.
200
n=18
r=0.57
p=0.015
•
•
-.:... 100
~
•
E
;::J
(j)
en
.£
0
.,....
(.)
(.)
50
•
25
2
•
5
CC10 in BAL mg·f
10
20
30
1
Fig. 6. - Correlation between Clara cell protein (CClO) in serum and
in BAL in healthy nonsmokers (log scale). BAL: bronchoalveolar lavage.
CClO was present in serum of nonsmokers at
concentrations which were on average forty times
lower than in BAL (fig. 5). Serum CClO increased
significantly in sarcoidosis patients, whereas it decreased in patients with COPD. In the serum
from smokers and lung cancer patients, the protein
showed a tendency to decrease (p=0. 12 and 0.06, respectively).
In the BAL of healthy nonsmokers (n=25), the CClO
concentration was significantly correlated with that of
albumin (r==0.57, p=0.004) and total protein (r=0.6,
p=0.002, n=23). No correlation was observed with ~2microglobulin (r=0.34, p=O.l), the secretory component
(r=0.27, p=0.19), cell counts (r=0.20, p=0.41), or age
(r=0.06, p=0.8). Interestingly, a significant correlation
was found between CClO in serum and in BAL of
healthy nonsmokers (fig. 6).
When other groups were considered separately,
the only noteworthy correlation which emerged was
that between BAL lymphocyte count and CClO in the
serum of patients with sarcoidosis (n=l2, r=0.72,
1236
A. BERNARD ET AL.
p=0.008). The concentration of CC10 in BAL (absolute values or expressed as protein/albumin ratio)
did not vary with the stage of sarcoidosis (analysis of
variance (ANOVA), F=l.1 and 0.6 with p=0.6 and
0.57, respectively). No statistically significant
association could be established with the smoking
history, although smokers with a lifetime smoking
>10 pack-years had on average a lower CC10 level
in BAL (1.38 mg·/· 1, n=6) than those with a lifetime smoking <10 pack-years (2.23 mg·/· 1, n=8)
(p=0.22).
Discussion
Data reported here show that CClO is one of the
most abundant proteins produced locally in the respiratory tract, as assessed by BAL measurements. In
BAL from healthy nonsmokers, its concentration
averaged 19% of that of albumin and 7% of that of
total protein. These estimates largely exceed those of
SJNGH and eo-workers (7), who reported that CC10
represented only 0.15% of the total protein content of
BAL. A possible explanation for this discrepancy is
that SJNGH and eo-workers [7) had adsorbed their antiserum with human serum and perhaps reduced its
activity by CC10 present in serum.
CC10 was originally reported to be secreted exclusively by the Clara cells [4). More recently, using
molecular biology techniques, BROERS et al. (17) demonstrated that in the normal lung the CC10 gene was
expressed in nonciliated columnar cells in large and
small bronchi as well as in bronchioles. These cells
probably contribute to most of the CClO secreted in
the respiratory tract. Since CC10 levels in BAL are
a factor of 40 greater than that in plasma (without taking into account the dilution by the BAL procedure),
the possibility of a plasma transudation can be virtually excluded, unless to invoke a hypothetical and
improbable intravascular secretion of CC10 followed
by active transport from plasma into the respiratory
tract. To ascertain that changes in a BAL soluble
component are not artificially generated by variations
in the amount of epithelial lining fluid recovered, it
is a common practice to express the concentrations
relatively to that of total protein or albumin. However, as albumin (like most BAL proteins) and CC10
have different origins, this mode of expression may be
misleading. In diseases associated w ith a defective
alveolar capillary barrier, resulting in an increased transudation of plasma proteins (e.g. sarcoidosis) [18], the
CC10/albumin ratio may be decreased despite normal
values of CC10. It is even questionable whether this
mode of expression corrects the values of CC10 for
the variable dilution of the BAL sample, since it does
not reduce the dispersion of values. Nevertheless, we
have checked that the reduction of CClO in the BAL
of smokers and in the affected lung side of cancer patients was still statistically significant after adjustment
for the albumin levels (which was in fact expected
since the latter were not decreased in these patients).
The possibility of an artificial reduction of CClO, due
to an incomplete recovery during the lavage procedure,
being formally ruled out, we will refer only to absolute values in the remainder of the discussion.
Age does not seem to influence the concentration of
CC10 in BAL, at least in the age range of nonsmokers examined here (21-65 yrs). This is important for
interpreting the data of the present study, since we
could not match cancer patients and controls for age.
CClO in BAL also seems to be independent of sex,
in contrast to its urinary excretion, which is significantly higher in males than females. As shown
elsewhere [9, 10], the sex dependency of urinary CClO
is due to a secretion of the protein by the male urogenital tract. By contrast, CClO in BAL appears to
be very sensitive to tobacco smoking. A reduction of
53% on average was found in a group of current
smokers with normal X-rays and lung function tests.
This is not related to a global reduction of the BAL
protein content caused by smoking, since no difference
was seen in the concentrations of albumin, ~ 2 microglobulin or the secretory component. No statistically significant association could be established with
the smoking history, presumably because of the low
number of subjects and also differences in smoking
habits (e.g. degree of smoke inhalation). It is noteworthy, however, that smokers with a smoking history
higher than 10 pack-years had CC10 levels in BAL on
average 40% lower than smokers with a smoking history below 10 pack-years. A further decrease of CC10
was found in the serum and in BAL of COPD patients
compared to smokers, which supports the hypothesis
of a dose-dependent reduction of CClO in the respiratory tract as a result of tobacco smoking. Another
line of evidence comes from a recent investigation
showing that CClO in the serum of smokers is inversely related to the pack-year smoking history [19).
Since CClO in lung lavage is unlikely to derive
from serum, it seems reasonable to assume that the
reduction of its content in BAL from smokers and
patients with COPD reflects a reduced synthesis and/
or release of the protein by Clara cells. Experimental
evidence is now accumulating that the Clara cell is a
very sensitive target for a number of pneumotaxic
chemicals. This vulnerability of the Clara cell
probably stems from its high potential for metabolizing
xenobiotics, namely via cytochrome P-450 dependent
mono-oxygenases [20). Many polycyclic hydrocarbons, of which cigarette smoke contains hundreds,
require activation by the cytochrome P-450 to produce
reactive metabolites. An attractive hypothesis would
be that these toxic species progressively destroy Clara
cells and thereby decrease the production of CC10 in
the respiratory tract. This hypothesis agrees with the
observation of LuMSDEN et al. [21], showing a reduction in Clara cell number in the distal airways of
smokers. If, as postulated [7, 8), CC10 is a natural
immuno-suppressor down-regulating the immune system, its diminution in the respiratory tract of smokers
could explain some inflammatory changes induced by
smoking.
1237
CIARA CELL PROTEIN IN SERUM AND BAL
Pulmonary sarcoidosis does not affect the concentration of CClO in BAL fluids, despite significantly
higher levels of albumin, ~ 2 -microglobulin and of the
secretory component (table 1). As sarcoidosis is
known to preferentially involve the lung parenchyma
rather than the bronchial tree, this reinforces the
specificity of CC10 as an indicator of Clara cell
damage or dysfunction. In BAL of patients with lung
cancer, CC10 undergoes changes which are also
distinct from those affecting albumin, ~ 2 -microglobulin
or the secretory component. In the BAL from
the affected side, CClO was significantly reduced
compared to values in the unaffected side and values
in nonsmokers, whereas other proteins were unchanged. The lack of statistically significant reduction of CC10 in the uninvolved side compared to
nonsmokers is the consequence to two outlying
values. After exclusion of the latter, the geometric
mean falls to 1.9 mg·/·1 and becomes significantly different from that of healthy nonsmokers, as expected,
since most of these cancer patients were smokers. At
the present time, no satisfactory explanation can be
proposed for these outlying values. The hypothesis
could be raised that some lung cancer cells secrete
CC10 but this was not supported by a recent study
[17].
Another interesting finding in the present study is
that the secretion of CC10 in the respiratory tract
might, in some conditions, be assessed indirectly by
measuring the protein in serum. The concentrations of
CClO in serum and BAL of healthy nonsmokers are
correlated and the reduction of CC10 observed in BAL
of smokers and patients with COPD or lung cancer is,
on average, mirrored by the levels in serum. All evidence presently available indicates that the small
amounts of CC10 occurring in serum derive exclusively from the respiratory tract. Studies [4, 5) which
have investigated the distribution of CC10 in animals
have failed to detect the protein other than in the Clara
cells of the pulmonary epithelium. In agreement with
these studies, we also found high concentrations
of CClO in human lung parenchyma (around 6 ppm),
but we could not detect the protein in homogenates of
liver and kidney, or in prostate or epididymis, where
uteroglobulin the equivalent of CClO in rabbit (6]
is present. It is unlikely that the sex-dependent
postrenal secretion of CC10, that we have recently
observed, exerts any influence on the serum levels of
CClO since no difference between men and women
could be established for this parameter [10). At the
present time, the only known limitation of the serum
CClO is that, owing to its small size, the protein is
rapidly eliminated by glomerular filtration and, as a
corollary, rises markedly during renal insufficiency
[10]. If one excludes situations associated with reduced renal function, CC10 in serum might be an indicator of the total amount of the protein synthetized
by the epithelial cells of the respiratory tract, which
would be more reliable than CClO in BAL, since the
latter is influenced by the variable dilution of the
sample.
At least two mechanisms might, in theory, account
for the passage of CClO from the respiratory tract into
the blood. The first is a passive transudation of the
protein across the pulmonary epithelium similar, but
in the opposite direction, to that of albumin and ~2 microglobulin. This diffusion probably forms the
basis of the correlation between CClO in BAL and serum observed in healthy nonsmokers. In pathological
conditions, however, the barrier between the surface of
respiratory epithelium and the vascular compartment
may be disrupted, upsetting the diffusional equilibrium
between CC10 in serum and in the respiratory tract.
This second mechanism might explain the elevation of
CC10 in the serum of patients with sarcoidosis. The
existence of an enhanced passage of proteins across
the blood/bronchoalveolar space barriers in sarcoidosis is supported by the significant elevation of albumin, ~ 2-microglobulin (table 1) and other plasma
proteins in BAL fluids [18]. The correlation between
CC10 in serum (but not in BAL) and the BAL
lymphocyte count is also interesting because they
might both represent indices of local immune reaction
in sarcoidosis. The value of CClO measurements in
serum compared to other markers, such as angiotensin
converting enzyme and lysozyme, at diagnosis and
during the course of pulmonary sarcoidosis remains to
be evaluated.
In conclusion, the present study suggests that CC10
in BAL and also in serum might serve as a sensitive
and specific marker of a dysfunction or damage of
nonciliated bronchial cells and in particular of Clara
cells. As the latter appear very vulnerable to toxic injury, the assay of CC10 in serum might be a useful
test for monitoring populations exposed to bronchial
toxins.
Acknowledqements: We gratefully acknowledge
X. Dumont and J.P. Dehennin for their expert technical assistance. This study was supporled by the
Fonds de la Recherche Scientifique Mtdicalc (Bel·
gium) and the Commission of the European Commu·
nities. A.B. is Maitre de rccherches du Fonds Na·
tional de la Recherche Scientifique.
References
1. Reynolds HY, Newball HH. - Analysis of proteins
and respiratory cell from human lungs by bronchial lavage.
J Clin Med, 1974; 84: 559-573.
2. Merrill WW, Goodenberger D, Strober W, et al. Free secretory component and other proteins in human lung
lavage. Am Rev Respir Dis, 1980; 122: 156- 161.
3. Burnett D. - lmmunoglobulins in the lung. Thorax,
1986; 41: 337-344.
4. Singh G, Katyal SL, Ward JM, et al. - Secretory proteins of the lung in rodents: immunocytochemistry. J
Histochem Cytochem, 1985; 33: 564-570.
5. Singh G, Katyal SL, Brown WE, et al. - Amino acid
and cDNA nucleotid sequences of human clara cell 10 kDa
protein. Biochim Biophys Acta, 1988; 950: 329-337.
6. Singh G, Katyal SL, Brown WE, et al. - Clara cell
10 kDa protein (CC10). Comparison of structure and
1238
A. BERNARD ET AL.
function to uteroglobulin. Biochim Biophys Acta, 1990;
1039: 348-355.
7. Singh G, Singh J, Katyal SL, et al. - Identification,
cellular localization, isolation and characterization of human
Clara cell-specific 10 kD protein. J Histochem Cytochem,
1988; 1: 73-80.
8. Jackson PJ, Turner R, Keen JN, Booksbank RA,
Cooper EH. - Purification and partial amino acid sequence
of human urine protein 1. J Chromatography, 1988; 452:
359-367.
9. Bernard A, Roels H, Lauwerys R, et al. - Protein 1
is a secretory protein of the respiratory and urogenital tracts
identical to the Clara cell protein. Clin Chem, 1992; 38:
434-435.
10. Bernard A, Roels H, Lauwerys R, et al. - Human
urinary protein 1: evidence for identity with the Clara cell
protein and occurrence in respiratory tract and urogenital
secretions. C/in Chim Acta, 1992; 207: 239-249.
11. Bernard A, Lauwerys R, Noel A, Vandeleene B,
Lambert A. - Determination of protein 1 in normal
and pathological urine. Clin Chim Acta, 1991; 201: 231246.
12. Sobin LH, Hermanek P, Hutter RVP. - TNMclassification of malignant tumours. 4th ed, UICC Geneva,
1987.
13. Standardization of lung function tests. European Community for Coal and Steel, Luxembourg, 1971.
14. Marchandise FXP, Mathieu B, Francis C, Sibille Y.
- Local increase of antiprotease and neutrophil elastasea1-proteinase inhibitor complexes in lung cancer. Eur Resp
], 1989; 2: 623-629.
15. Bernard A, Lauwerys R. - Continuous flow system
for the automation of latex immunoassay by particle counting. Clin Chem, 1983; 30: 1007-1011.
16. Fedman D, Cagnon J, Hofmann R, Simpson J. - Abacus Concepts, Inc., 1988.
17. Broers JLV, Jensen SM, Travis WD, et al. - Expression of surfactant-associated protein A and Clara cell 10
Kilodalton mRNA in neoplastic and non-neoplastic human
lung tissue as detected by in situ hybridization. Lab Invest,
1992; 66: 337-346.
18. Delacroix DL, Marchandise FX, Francis C, Sibille Y.
- Alpha2 -macroglobulin, monomeric and polymeric immu·
noglobulin A, and immunoglobulin M in bronchoalveolar
lavage. Am Rev Respir Dis, 1985; 132: 829-835.
19. Bernard A, Roels H, Buchet JP, Lauwerys R.
- Decrease of Clara cell protein in the serum of smokers.
(Letter) Lancet, 1992; 339: 1620.
Clara
20. Richards RJ, Oreffo VIC, Lewis RW.
cell cultures from the mouse and their reaction to
bronchial toxins. Environ Health Perspect, 1990; 85: 119127.
21. Lumsden AB, McLean A, Lamb D. - Goblet and
Clara cells of human distal airways: evidence for smoking
induced changes in their numbers. Thorax, 1984; 39:
844-849.